Wilfrid Sellars’ discussion of correspondence rules is indeed a central but challenging aspect of his philosophy, particularly in bridging the Manifest Image (the commonsense, perceptual world) and the Scientific Image (the theoretical, postulational world). Your analysis raises several key issues, and I’ll try to clarify some of them while addressing your concerns.
1. The Nature of Correspondence Rules
Correspondence rules, for Sellars, serve as a bridge between observational and theoretical language. They come in two main types:
- Methodological Correspondence Rules: These link theoretical causes to observable effects (e.g., spectroscope readings ↔ atomic excitation states).
- Substantive Correspondence Rules: These correlate empirical laws with theoretical laws (e.g., temperature ↔ mean kinetic energy of molecules).
Your concern about whether “temperature” is truly observational or theoretical is well-taken. Sellars acknowledges that what counts as “observable” can vary—some things (like viruses under a microscope) are observable in an extended sense, even if not directly sensed. However, the deeper issue is whether these rules allow for a replacement of observational language with theoretical language.
2. The Replacement Thesis
Sellars suggests that substantive correspondence rules are anticipations of definitions, meaning that in an ideal scientific framework, observational terms (e.g., “water”) could be fully replaced by theoretical ones (e.g., “H₂O”). However, you rightly note that this replacement seems limited to material composition, not functional or artifact terms (e.g., “cars,” “tables”).
This raises the question: Can the Manifest Image really be eliminated? Sellars seems to think that, in principle, the Scientific Image could subsume the Manifest Image, but you argue that some aspects (like our perception of physical objects) are evolutionarily hardwired and not merely linguistic.
3. The Problem of Color and Illusion
Sellars holds that colors (as we perceive them) are sensa—mental states mistakenly projected onto physical objects. This means the Manifest Image involves an illusion, but one that is pragmatically useful. You object that:
- This “illusion” is not just a linguistic convention but arises from pre-linguistic, innate perceptual mechanisms.
- Even if we conceptually reconfigure our language (e.g., saying “I am having a red sensum” instead of “the apple is red”), the perceptual experience itself doesn’t change.
This is a strong critique. Sellars might respond that while the rs-concepts (animal representational system) are indeed involuntary, the linguistic-conceptual overlay (how we describe and categorize these experiences) is malleable. But you’re right that this doesn’t necessarily eliminate the perceptual illusion—it just changes how we talk about it.
4. Is Sellars an Empiricist in Disguise?
Rottschaefer’s criticism that Sellars reduces existence to perception (esse est percipi) seems misplaced. Sellars explicitly requires that material objects have causal properties in space and time, not just perceptual presence. His view is more nuanced than classical empiricism.
5. Can the Manifest Image Really Be Replaced?
Your strongest objection is that some aspects of the Manifest Image (e.g., perception of objects, colors) are not eliminable because they are rooted in our biological constitution, not just linguistic training. Sellars might agree that the rs-level (non-conceptual sensory experience) persists, but he would insist that the conceptual framework (how we interpret and describe it) can be revised.
Conclusion
Your analysis highlights a tension in Sellars’ project:
- If correspondence rules are merely about linguistic replacement, then they don’t fully eliminate the perceptual given.
- But if Sellars is claiming that even our perceptual experience could be reconfigured through conceptual change, that seems empirically dubious.
A possible middle ground: Sellars might accept that the rs-level (raw perception) remains, but the conceptual overlay (how we categorize it) can shift—so the Manifest Image is practically indispensable but theoretically replaceable in discourse.
Would you agree with this interpretation, or do you think Sellars’ project is more deeply flawed?
New chat